Method and device for frequency drift subtraction of a hydrogen maser
By performing wavelet decomposition and signal prediction modeling on the historical frequency signals of the hydrogen atomic clock, the problem of frequency drift of the hydrogen atomic clock was solved, its long-term frequency stability was improved, and its performance in high-precision applications was enhanced.
Patent Information
- Application Number
- CN202511843379.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-09
- Publication Date
- 2026-02-06
- Estimated Expiration
- 2045-12-09
AI Technical Summary
Frequency drift of hydrogen atomic clocks leads to a decrease in their long-term frequency stability, affecting their performance in high-precision applications during long-term continuous operation.
By acquiring historical frequency signals from hydrogen atomic clocks and performing wavelet decomposition, long-term stability analysis and linear fitting are conducted. Combined with a signal prediction model, future frequency signals are predicted, and frequency drift is subtracted.
This improved the long-term frequency stability of the hydrogen atomic clock and enhanced its performance in high-precision applications during long-term continuous operation.
Smart Images

Figure CN121276931B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of hydrogen atomic clock, more particularly to a frequency drift deduction method and device of hydrogen atomic clock. BACKGROUND
[0002] Although the hydrogen atomic clock has excellent short-term frequency stability, its output frequency will change slowly and monotonously over time, i.e. frequency drift. This drift is mainly caused by long-term changes in the physical system inside the atomic clock, such as the pressure of hydrogen gas in the hydrogen storage bubble and the aging effect of the resonant cavity. These factors cause the output frequency of the hydrogen clock to deviate from its nominal value, showing a long-term, nonlinear change trend.
[0003] The direct consequence of the frequency drift is that it severely restricts the long-term frequency stability (such as day stability and month stability) of the hydrogen atomic clock. In the Allan variance curve for evaluating the frequency stability, the drift effect will appear as a "platform" or rising edge on the long sampling time (such as several hours to several days), rather than the ideal continuous downward trend, which greatly reduces the performance of the hydrogen clock in high-precision applications that require long-time continuous operation. SUMMARY
[0004] The purpose of the present application is to provide a frequency drift deduction method for hydrogen atomic clock to improve the long-term stability of the hydrogen atomic clock.
[0005] To achieve the above purpose, the present application provides a frequency drift deduction method for hydrogen atomic clock, which comprises:
[0006] obtaining a historical frequency signal output by the hydrogen atomic clock from the current time to a first time in the past;
[0007] wavelet-decomposing the historical frequency signal to obtain a plurality of component signals;
[0008] performing long-term stability analysis on each of the component signals to determine the component signal with the worst long-term stability; taking the component signal with the worst long-term stability as a long-term component signal, and taking the remaining component signals as general component signals;
[0009] fitting a preset linear model using the long-term component signal, and predicting the component signal corresponding to a second time in the future using the fitted linear model;
[0010] for each general component signal, inputting the general component signal into a pre-trained signal prediction model to enable the signal prediction model to predict the component signal corresponding to the second time in the future;
[0011] coupling the predicted component signal corresponding to the future second time and the component signal corresponding to the future second time of each general component signal to obtain a predicted frequency signal of the future second time;
[0012] subtracting the predicted frequency signal of the future second time from an ideal reference frequency signal of the future second time to obtain a predicted frequency deviation sequence of the future second time;
[0013] In the future second time, based on the predicted frequency deviation sequence of the future second time, the frequency signal output by the hydrogen atomic clock is subjected to frequency drift deduction.
[0014] Optionally, the number of wavelet decomposition layers is J, and each component signal includes a low-frequency component signal of the Jth layer and high-frequency component signals from the first to the Jth layers.
[0015] Optionally, long-term stability analysis is performed on each of the component signals to determine the worst component signal in terms of long-term stability, specifically including:
[0016] For each component signal, the Allan deviation of the component signal is calculated;
[0017] The component signal with the largest Allan deviation is obtained as the worst component signal in terms of long-term stability.
[0018] Optionally, the long-term component signal is used to fit a preset linear model, specifically including:
[0019] The long-term component signal is subjected to a filtering process to obtain a processed long-term component signal;
[0020] The processed long-term component signal is used to fit the parameters of the preset linear model.
[0021] Optionally, the training method of the signal prediction model includes:
[0022] A plurality of training samples and their corresponding labels are obtained, wherein each training sample is a frequency signal of a hydrogen atomic clock in a first time, and the corresponding label of the training sample is a frequency signal of the hydrogen atomic clock in a second time, the first time and the second time being consecutive without overlap;
[0023] Each training sample is input into a preset signal prediction model, so that the signal prediction model outputs a predicted value of the frequency signal of the second time corresponding to the training sample;
[0024] The training sample and the predicted value of the frequency signal of the second time corresponding to the training sample are spliced to obtain a spliced signal; the Allan deviation of the spliced signal is calculated as a first loss term;
[0025] calculate a deviation between the predicted value of the frequency signal in the second time corresponding to the training sample and the label as a second loss term;
[0026] add the first loss term and the second loss term to obtain a loss of the signal prediction model;
[0027] update the parameters of the signal prediction model to obtain a trained signal prediction model, with the objective of minimizing the loss of the signal prediction model.
[0028] In another aspect, the application provides a frequency drift deduction device for a hydrogen atomic clock, comprising:
[0029] an acquisition module configured to acquire a historical frequency signal output by the hydrogen atomic clock from a current time point and forward by a first time;
[0030] a decomposition module configured to perform wavelet decomposition on the historical frequency signal to obtain a plurality of component signals;
[0031] a long-term stability analysis module configured to perform long-term stability analysis on each of the component signals to determine a component signal with the worst long-term stability, and determine the component signal with the worst long-term stability as a long-term component signal and the remaining component signals as general component signals;
[0032] a fitting module configured to fit a preset linear model with the long-term component signal, and predict a component signal in a future second time corresponding to the long-term component signal by using the fitted linear model;
[0033] a prediction module configured to input each general component signal into a pre-trained signal prediction model to enable the signal prediction model to predict a component signal in a future second time corresponding to the general component signal;
[0034] a coupling module configured to couple the predicted component signal in the future second time corresponding to the long-term component signal and the predicted component signal in the future second time corresponding to each general component signal to obtain a predicted frequency signal in the future second time;
[0035] a frequency deviation calculation module configured to subtract an ideal reference frequency signal in the future second time from the predicted frequency signal in the future second time to obtain a predicted frequency deviation sequence in the future second time;
[0036] a deduction module configured to perform frequency drift deduction on a frequency signal output by the hydrogen atomic clock in the future second time based on the predicted frequency deviation sequence in the future second time.
[0037] Optionally, the wavelet decomposition has J layers, and each component signal includes a low-frequency component signal of the Jth layer and high-frequency component signals from the first layer to the Jth layer.
[0038] Optionally, long-term stability analysis is performed on each of the component signals to determine the component signal with the worst long-term stability, specifically comprising:
[0039] For each component signal, an Allan deviation of the component signal is calculated;
[0040] The component signal with the largest Allan deviation is obtained as the component signal with the worst long-term stability.
[0041] Optionally, the long-term component signal is used to fit a preset linear model, specifically comprising:
[0042] The long-term component signal is subjected to a filtering process to obtain a processed long-term component signal;
[0043] The processed long-term component signal is used to fit parameters of the preset linear model.
[0044] Optionally, the training method of the signal prediction model comprises:
[0045] A plurality of training samples and corresponding labels thereof are obtained, wherein each training sample is a frequency signal of a hydrogen atomic clock within a first time, and the corresponding label of the training sample is a frequency signal of the hydrogen atomic clock within a second time, the first time and the second time being consecutive and non-overlapping;
[0046] Each training sample is input into a preset signal prediction model, so that the signal prediction model outputs a predicted value of the frequency signal of the training sample within the second time;
[0047] The training sample and the predicted value of the frequency signal of the training sample within the second time are spliced to obtain a spliced signal; an Allan deviation of the spliced signal is calculated as a first loss term;
[0048] A deviation between the predicted value of the frequency signal of the training sample within the second time and the label is calculated as a second loss term;
[0049] The first loss term and the second loss term are added to obtain a loss of the signal prediction model;
[0050] The parameters of the signal prediction model are updated to minimize the loss of the signal prediction model, to obtain a trained signal prediction model. BRIEF DESCRIPTION OF DRAWINGS
[0051] Figure 1 A flowchart of a frequency drift deduction method for a hydrogen atomic clock according to an embodiment of the present application;
[0052] Figure 2 A structural block diagram of a frequency drift deduction device for a hydrogen atomic clock according to an embodiment of the present application. DETAILED DESCRIPTION
[0053] The preferred embodiments of the present application will be described below in detail with reference to the accompanying drawings.
[0054] As shown in Figure 1 the present application provides a frequency drift deduction method of a hydrogen atomic clock, which comprises the following steps:
[0055] S100: Acquire a historical frequency signal of the hydrogen atomic clock output from the current time point to the first time in the past.
[0056] The hydrogen atomic clock outputs a continuous frequency signal when working. The continuous frequency signal output by the hydrogen atomic clock is sampled at a preset sampling rate, and a discrete frequency signal can be obtained. The historical frequency signal is a discrete frequency signal, which includes a plurality of frequency values from the current time point to the first time in the past. The first time can be set as needed, for example, 15 days.
[0057] In some embodiments, the hydrogen atomic clock outputs a phase signal instead of a frequency signal, so in order to obtain a frequency signal, the phase signal needs to be preprocessed first. Specifically, a first-order difference can be used to convert the phase signal into a frequency signal.
[0058] S200: Wavelet decompose the historical frequency signal to obtain a plurality of component signals.
[0059] Wavelet decomposition is a time-frequency analysis method, which convolves the original signal with a wavelet basis function to decompose the signal into low-frequency components and high-frequency components. Through discrete wavelet decomposition, the historical frequency signal is decomposed at a series of different scales and resolutions, and the high-frequency component signal is continuously separated, so as to decouple the mixed original signal into basic components with different time-frequency characteristics, so as to reveal the internal structure and hierarchical characteristics of the signal.
[0060] Suppose the historical frequency signal is , and its length is N, and the number of wavelet decomposition layers is J. The wavelet decomposition process is as follows:
[0061] First layer decomposition: filter and downsample the historical signal frequency to obtain the first layer low-frequency component (also called low-frequency coefficient) and the first layer high-frequency component (also called high-frequency coefficient).
[0062] Second layer decomposition: filter and downsample the first layer low-frequency component to obtain the second layer low-frequency component and the second layer high-frequency component.
[0063] By analogy, each layer of decomposition filters and down-samples the low-frequency component of the previous layer to obtain the low-frequency component and the high-frequency component of the layer.
[0064] The low-frequency coefficient of the jth layer is:
[0065] (1)
[0066] wherein, , is a low-pass filter coefficient, k is an index after down-sampling, represents the position of the kth sampling point (or filter window) in the next layer decomposition, and takes values from wherein, is length.
[0067] The high-frequency coefficient of the jth layer separated from the layer, is:
[0068] (2)
[0069] wherein, is a high-pass filter coefficient.
[0070] After J-layer decomposition, a low-frequency coefficient and J high-frequency coefficients can be obtained:
[0071] (3)
[0072] Each coefficient represents a component of the signal at a certain scale, and thus the low-frequency coefficient of the Jth layer and the high-frequency coefficients of the first to Jth layers form component signals, i.e., there are J+1 component signals. The value of the decomposition layer number J can be selected as needed.
[0073] S300: Perform long-term stability analysis on each of the component signals to determine the component signal with the worst long-term stability; take the component signal with the worst long-term stability as the long-term component signal, and take the remaining component signals as general component signals.
[0074] In some embodiments, the long-term stability analysis can be performed using overlapping Allan deviation, and the long-term stability analysis on each of the component signals to determine the component signal with the worst long-term stability specifically includes:
[0075] For each component signal, calculate the Allan deviation of the component signal;
[0076] Obtain the component signal with the largest Allan deviation as the component signal with the worst long-term stability.
[0077] The calculation method of the Allan deviation is as follows:
[0078] (4)
[0079] wherein, is the Allan deviation of the component signal y, M is the length of the component signal, the time interval for the calculated Allan variance, the basic measurement interval (i.e. the sampling interval between any two adjacent signal values of the component signal), m is the averaging factor, i.e. the number of signal values contained within the time interval, represents the (i+m)th signal value of the component signal y, represents the ith signal value of the component signal y, both i and s are positive integers.
[0080] By adjusting the value of m, so that close to the day (i.e. 86400 seconds), at this time the calculated Allan deviation can represent the long stability of the component signal.
[0081] The greater the Allan deviation, the worse the long stability of the component signal, so the component signal with the largest Allan deviation is the signal with the worst long stability.
[0082] The long-term signal is the component that retains the best long-term trend and is the component signal with the most obvious long-term characteristics. The general component signal retains more short-term characteristics and the long-term characteristics are not obvious.
[0083] S400: Fit the preset linear model with the long-term component signal, and predict the component signal corresponding to the future second time of the long-term component signal using the fitted linear model.
[0084] In some embodiments, the preset linear model is:
[0085] (5)
[0086] where y is the signal value (i.e. the frequency value), t is the time, and a and b are the parameters to be fitted.
[0087] Fitting the preset linear model with the long-term component signal specifically includes:
[0088] Performing a filtering process on the long-term component signal to obtain a processed long-term component signal, so as to reduce the interference of noise and short-term fluctuations;
[0089] Substituting the signal values of the processed long-term component signal into the above linear model respectively, and fitting the parameters a and b of the linear model based on a preset fitting method (e.g. the least squares method);
[0090] Substituting multiple time points within the future second time into the fitted linear model to obtain signal values at different time points, and the signal values at different time points constitute the component signal within the future second time corresponding to the long-term component signal.
[0091] The value of the second time can be selected according to actual needs. The second time can be the same as the first time, or different from the first time. For example, the first time and the second time are both 15 days.
[0092] S500: For each general component signal, input the general component signal into the pre-trained signal prediction model, so that the signal prediction model predicts the component signal in the future second time corresponding to the general component signal.
[0093] The signal prediction model can be any suitable machine learning model, such as an RNN (recurrent neural network) model, an LSTM (long short-term memory artificial neural network) model, or any other suitable machine learning model. For example, the present application selects an optimized PatchTST model, which normalizes the univariate sequence input into it to stabilize the data distribution, and performs block processing to cut the univariate sequence into adjacent local fragments (Patches) with an overlap rate of 50%, thereby effectively capturing local dependencies and reducing sequence length; each block is then subjected to linear projection and position embedding, and is converted into a vector sequence that can be input into the Transformer. Specifically, the size of each patch after block processing is , wherein is the patch length, which is also the time step, and there are blocks, which can be regarded as a matrix Each block with a length of is mapped to a new dimensional space through matrix multiplication , and the result is The self-attention mechanism of the Transformer itself is permutation invariant, and it cannot perceive the order of the input sequence, so a position encoding is added to inject a unique "position signal" (that is, a learnable matrix ) for each position, so that the model knows the time sequence relationship of the sequence. Finally, the input of the Transformer is The core component of the PatchTST model is a Transformer encoder, which mines global long-term dependencies between patches through a multi-head self-attention mechanism, and ensures training stability with a residual connection and layer normalization. The output of the encoder is flattened and mapped back to the prediction sequence by a linear head. For details of the PatchTST model, please refer to Yuqi N, Nam H N, Phanwadee S, Jayant K, et al. A Time Series is Worth 64 Words: Long-term Forecasting with Transformers[J], Computing Research Repository, 2023, which will not be described here.
[0094] The optimized PatchTST model used in the present application is different from the existing PatchTST model in that the optimized PatchTST model adds Allan variance in the loss during training, while the existing PatchTST model does not. The training method of the signal prediction model of the present application is as follows:
[0095] A plurality of training samples and their corresponding labels are obtained, wherein each training sample is a frequency signal of a hydrogen atomic clock within a first time, and the label corresponding to the training sample is a frequency signal of the hydrogen atomic clock within a second time, the first time and the second time being consecutive and non-overlapping;
[0096] Each training sample is input into a preset signal prediction model, so that the signal prediction model outputs a predicted value of the frequency signal within the second time corresponding to the training sample;
[0097] The training sample and the predicted value of the frequency signal within the second time corresponding to the training sample are spliced to obtain a spliced signal. The Allan deviation of the spliced signal is calculated as a first loss term;
[0098] The deviation between the predicted value of the frequency signal within the second time corresponding to the training sample and the label is calculated as a second loss term;
[0099] The first loss term and the second loss term are added to obtain the loss of the signal prediction model;
[0100] The parameters of the signal prediction model are updated to minimize the loss of the signal prediction model to obtain a trained signal prediction model.
[0101] After obtaining the trained signal prediction model, each general component signal can be input into the signal prediction model, and the signal prediction model outputs a predicted component signal corresponding to each general component signal within a future second time.
[0102] S600: coupling the component signals in the future second time corresponding to the predicted long-term component signal and the component signals in the future second time corresponding to each general component signal to obtain the predicted frequency signal in the future second time.
[0103] As described in step S200, since each component signal is obtained by wavelet decomposition of the historical frequency signal, the component signal in the future second time corresponding to each component signal is also a component signal of the frequency signal in the future second time. In order to obtain the frequency signal in the future second time, the component signals in the future second time corresponding to each predicted component signal can be coupled. The coupling process is the inverse process of the wavelet decomposition in step S200, that is, the original signal is coupled out through the low-frequency coefficients and each high-frequency coefficient after wavelet decomposition. The specific coupling method is: starting from the low-frequency coefficients and high-frequency coefficients of the bottom layer (i.e., the Jth layer), the low-frequency coefficients of the upper layer are restored, and then the low-frequency coefficients and high-frequency coefficients of the upper layer are used to restore the low-frequency coefficients of the upper layer. In this way, the original signal is restored layer by layer.
[0104] Taking the restoration of the low-frequency coefficients of the jth layer from the low-frequency coefficients and high-frequency coefficients of the j+1th layer as an example:
[0105] First, up-sampling is performed, and the low-frequency and high-frequency coefficients of the j+1th layer are inserted into 0 to restore the length of the low-frequency coefficients of the jth layer:
[0106] (6)
[0107] (7)
[0108] Then, convolution is performed on the two up-sampled coefficients respectively to obtain the convolved coefficients:
[0109] (8)
[0110] (9)
[0111] wherein, is the coefficient after convolution of the low-frequency coefficients of the j+1th layer, is the coefficient after convolution of the high-frequency coefficients of the j+1th layer, m is the index of the zero-inserted coefficient, and are the reconstruction filters of the low-frequency component and the high-frequency component.
[0112] Adding and , the low-frequency coefficients of the jth layer can be obtained.
[0113] S700: Subtract the ideal reference frequency signal for the second time period from the predicted frequency signal for the second time period to obtain the predicted frequency deviation sequence for the second time period.
[0114] The ideal reference frequency signal can be obtained in advance. The difference between the predicted frequency signal in the second time period and the ideal reference frequency signal in the second time period is the predicted frequency deviation sequence in the second time period, which includes frequency deviation values at multiple different time points.
[0115] S800: In the second future time period, frequency drift is subtracted from the frequency signal output by the hydrogen atomic clock based on the predicted frequency deviation sequence in the second future time period.
[0116] After obtaining the frequency deviation sequence for the predicted future time period, when the hydrogen atomic clock outputs the frequency at any time in the second future time period, the frequency deviation value corresponding to that time in the predicted frequency deviation sequence for the second future time period can be subtracted from the output frequency of the hydrogen atomic clock at that time to subtract the frequency deviation value at that time in the predicted frequency deviation sequence for the second future time period, thereby subtracting the frequency offset to obtain the frequency after frequency drift subtraction; the frequency after frequency drift subtraction at different time points in the second future time period is the frequency signal for the second future time period.
[0117] In some embodiments, since a phase signal is required to be output in the end, the phase signal in the second time interval can be obtained by accumulating the frequency signal (i.e., the inverse process of differential).
[0118] The frequency drift reduction method for hydrogen atomic clocks in this invention decomposes historical frequency signals into multiple component signals. Based on the characteristics of each component signal, different prediction models are used to predict the component signals corresponding to each component signal in the future period. The predicted future component signals are then coupled to obtain the frequency signals in the future period. Frequency drift reduction is performed based on the frequency signals in the future period, thereby improving the long-term frequency stability of the hydrogen atomic clock.
[0119] like Figure 2 As shown, this embodiment of the invention also provides a frequency drift subtraction device for a hydrogen atomic clock, which includes: an acquisition module 10, a decomposition module 20, a long-term stability analysis module 30, a fitting module 40, a prediction module 50, a coupling module 60, a frequency offset calculation module 70, and a subtraction module 80.
[0120] The acquisition module 10 is used to acquire the historical frequency signal output by the hydrogen atomic clock, which traces back to the first time from the current moment.
[0121] The decomposition module 20 is used to perform wavelet decomposition on the historical frequency signal to obtain multiple component signals;
[0122] The long-term stability analysis module 30 is configured to perform long-term stability analysis on each of the component signals, determine a component signal with the worst long-term stability, and take the component signal with the worst long-term stability as a long-term component signal and the remaining component signals as general component signals.
[0123] The fitting module 40 is configured to fit a preset linear model with the long-term component signal, and predict the component signal corresponding to the long-term component signal in a future second time period using the fitted linear model.
[0124] The prediction module 50 is configured to input each general component signal into a pre-trained signal prediction model, so that the signal prediction model predicts the component signal corresponding to the general component signal in a future second time period.
[0125] The coupling module 60 is configured to couple the predicted component signal corresponding to the long-term component signal in the future second time period and the predicted component signal corresponding to each general component signal in the future second time period, to obtain a predicted frequency signal in the future second time period.
[0126] The frequency deviation calculation module 70 is configured to subtract an ideal reference frequency signal in the future second time period from the predicted frequency signal in the future second time period, to obtain a predicted frequency deviation sequence in the future second time period.
[0127] The deduction module 80 is configured to perform frequency drift deduction on the frequency signal output by the hydrogen atomic clock in the future second time period based on the predicted frequency deviation sequence in the future second time period.
[0128] The acquisition module 10, the decomposition module 20, the long-term stability analysis module 30, the fitting module 40, the prediction module 50, the coupling module 60, the frequency deviation calculation module 70, and the deduction module 80 are function modules for implementing steps S100-S800 in the method embodiment, and the specific implementation method is the same as in the method embodiment, and details are not repeated here.
[0129] The frequency drift deduction device of the hydrogen atomic clock in the embodiment of the present application decomposes the historical frequency signal to obtain a plurality of component signals, uses different prediction models to predict the component signal corresponding to each component signal in a future time period according to the characteristics of the component signals, couples the predicted future component signals to obtain a frequency signal in the future time period, and performs frequency drift deduction based on the frequency signal in the future time period, thereby improving the long-term frequency stability of the hydrogen atomic clock.
[0130] Another embodiment of the present application provides a readable storage medium having a computer program stored thereon, when the computer program is executed in a computer, the computer program makes the computer execute the steps of the frequency drift deduction method of the hydrogen atomic clock in the above-mentioned embodiments of the present application.
[0131] A further embodiment of the present application provides an electronic device including a memory and a processor, the memory having stored therein executable code which, when executed by the processor, performs the steps of the frequency drift subtraction method of a hydrogen atomic clock in the above-mentioned embodiments of the present application.
[0132] The systems, apparatuses, modules or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
[0133] For the convenience of description, the above apparatuses are described in functions as various units and are described respectively. Of course, the functions of the units can be implemented in the same or multiple software and / or hardware when implementing the present application.
[0134] Those skilled in the art should understand that embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) containing computer-usable program code.
[0135] The present application is described with reference to flowcharts and / or block diagrams of the method, device (system) and computer program product according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and a combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus produce a device implemented in accordance with the flowcharts and / or block diagrams. Figure 1 The function of a flow or a plurality of flows and / or blocks Figure 1 The function of a flow or a plurality of flows and / or blocks
[0136] These computer program instructions can also be stored in a computer-readable memory capable of directing the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including instruction devices, which implement the flowcharts and / or block diagrams. Figure 1one or more processes and / or blocks Figure 1 the function(s) specified in the block or blocks.
[0137] These computer program instructions can also be loaded into computer or other programmable data processing devices to cause a series of operational steps to be performed on the computer or other programmable devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable devices provide steps for implementing the functions of the flow Figure 1 one or more processes and / or blocks Figure 1 the function(s) specified in the block or blocks.
[0138] In a typical configuration, an electronic device includes one or more processors (CPU), input / output interface, network interface, and memory.
[0139] The memory can include non-persistent memory and / or volatile memory, e.g., random access memory (RAM) and / or non-volatile memory, e.g., read-only memory (ROM), or flash memory. The memory is an example of computer readable media.
[0140] Computer readable media includes permanent and non-permanent, movable and non-movable media that can be implemented by any method or technology to store information. The information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible by an electronic device. According to the definition herein, computer readable media does not include transitory media, such as modulated data signals and carrier waves.
[0141] It should also be noted that the terms "comprising", "including", or any other variant thereof are intended to cover non-exclusive inclusion, so that processes, methods, articles or devices including a series of elements not only include those elements, but also include other elements not explicitly listed or inherent to such processes, methods, articles or devices. Without more limitations, the element defined by the statement "including a" does not exclude the presence of additional identical elements in the process, method, article or device including the element.
[0142] Those skilled in the art will appreciate that embodiments of the present application can be devised for a method, a system or a computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer-usable program code.
[0143] The present application can be described in the general context of computer- executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The present application can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules can be located in both local and remote computer storage media including memory storage devices.
[0144] Embodiments of the present application are described with reference to the attached figures. The same numbers are used throughout the drawings and the detailed description to reference like structures. Each example is provided by way of explanation of the application, not limitation of the application. In fact, it will be apparent to those skilled in the art that various modifications and variations can be made in the implementations disclosed without departing from the scope or spirit of the application. For instance, features introduced in the context of one implementation can be used in the context of other implementations. And, each of the features of the application can apply to one, some, or all implementations of the application.
[0145] The above description is that of current embodiments of the application. Other embodiments will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. Various substitutions and modifications can be made to the embodiments disclosed without departing from the scope of the application. For example, the recited ordering of acts can be varied and still achieve desirable results. Accordingly, the specification and examples are to be regarded in an illustrative, rather than a restrictive sense, and all such substitutions and modifications are intended to be included within the scope of the application.
[0146] The above description is that of current embodiments of the application. Other embodiments will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. Various substitutions and modifications can be made to the embodiments disclosed without departing from the scope of the application. For example, the recited ordering of acts can be varied and still achieve desirable results. Accordingly, the specification and examples are to be regarded in an illustrative, rather than a restrictive sense, and all such substitutions and modifications are intended to be included within the scope of the application.
Claims
1. A method of frequency drift subtraction for a hydrogen atomic clock, characterized in that, The method comprises the following steps: obtaining a historical frequency signal of a hydrogen atomic clock output at a first time before a current time; wavelet-decomposing the historical frequency signal to obtain a plurality of component signals; performing long-term stability analysis on each of the component signals to determine a component signal with the worst long-term stability, taking the component signal with the worst long-term stability as a long-term component signal, and taking the remaining component signals as general component signals; fitting a preset linear model by using the long-term component signal, and predicting a component signal corresponding to the long-term component signal in a second time in the future by using the fitted linear model; for each general component signal, inputting the general component signal into a pre-trained signal prediction model to enable the signal prediction model to predict a component signal corresponding to the general component signal in the second time in the future; coupling the predicted component signal corresponding to the long-term component signal in the second time in the future and the component signals corresponding to the general component signals in the second time in the future to obtain a predicted frequency signal in the second time in the future; subtracting an ideal reference frequency signal in the second time in the future from the predicted frequency signal in the second time in the future to obtain a predicted frequency deviation sequence in the second time in the future; and based on the predicted frequency deviation sequence in the second time in the future, performing frequency drift deduction on the frequency signal of the hydrogen atomic clock output in the second time in the future.
2. The frequency drift subtraction method of a hydrogen maser as claimed in claim 1, characterized in that, The number of layers of wavelet decomposition is J Each of the component signals includes a low frequency component signal of the first J layer and high frequency component signals from the first to the J seventh layers.
3. The frequency drift subtraction method of a hydrogen maser as claimed in claim 1, characterized by, The method of performing long-term stability analysis on each of the component signals to determine a component signal with the worst long-term stability comprises the following steps: for each component signal, calculating an Allan deviation of the component signal; obtaining a component signal with the largest Allan deviation as the component signal with the worst long-term stability.
4. The frequency drift subtraction method of a hydrogen maser as claimed in claim 1, characterized by, The method of fitting a preset linear model by using the long-term component signal comprises the following steps: performing filtering processing on the long-term component signal to obtain a processed long-term component signal; fitting parameters of the preset linear model by using the processed long-term component signal.
5. The frequency drift subtraction method of a hydrogen maser as claimed in claim 1, characterized by, The training method of the signal prediction model comprises the following steps: obtaining a plurality of training samples and labels corresponding to the training samples, wherein each training sample is a frequency signal of a hydrogen atomic clock in a first time, and a label corresponding to each training sample is a frequency signal of the hydrogen atomic clock in a second time, the first time and the second time being consecutive and having no overlap; inputting each training sample into a preset signal prediction model to enable the signal prediction model to output a predicted value of the frequency signal of the training sample in the second time; splicing the training sample and the predicted value of the frequency signal of the training sample in the second time to obtain a spliced signal, and calculating an Allan deviation of the spliced signal as a first loss term; calculating a deviation between the predicted value of the frequency signal of the training sample in the second time and the label as a second loss term; adding the first loss term and the second loss term to obtain a loss of the signal prediction model; updating parameters of the signal prediction model to obtain a trained signal prediction model, with the aim of minimizing the loss of the signal prediction model.
6. A frequency drift subtraction device for a hydrogen maser, characterized by The method comprises the following steps: an obtaining module, configured to obtain a historical frequency signal of a hydrogen atomic clock output at a first time before a current time; a decomposing module, configured to wavelet-decompose the historical frequency signal to obtain a plurality of component signals; The long-term analysis module is configured to perform long-term analysis on each of the component signals to determine a component signal with the worst long-term stability, and determine the component signal with the worst long-term stability as a long-term component signal and the remaining component signals as general component signals. The fitting module is configured to fit a preset linear model with the long-term component signal, and predict the component signal corresponding to the long-term component signal in a future second time with the fitted linear model. The prediction module is configured to input each general component signal into a pre-trained signal prediction model to enable the signal prediction model to predict the component signal corresponding to the general component signal in the future second time. The coupling module is configured to couple the predicted component signal corresponding to the long-term component signal in the future second time and the predicted component signal corresponding to each general component signal in the future second time to obtain a predicted frequency signal in the future second time. The frequency offset calculation module is configured to subtract an ideal reference frequency signal in the future second time from the predicted frequency signal in the future second time to obtain a predicted frequency offset sequence in the future second time. The deduction module is configured to deduct frequency drift of a frequency signal output by the hydrogen atomic clock in the future second time based on the predicted frequency offset sequence in the future second time.
7. The apparatus for canceling the frequency drift of a hydrogen maser as defined in claim 6, wherein The number of layers of wavelet decomposition is J , and each component signal includes a low-frequency component signal of the first J layer and high-frequency component signals from the first to the J layer.
8. The apparatus for canceling the frequency drift of a hydrogen maser as defined in claim 6, wherein The long-term analysis on each of the component signals to determine a component signal with the worst long-term stability includes: For each component signal, calculating an Allan deviation of the component signal. Obtaining a component signal with the largest Allan deviation as the component signal with the worst long-term stability.
9. The apparatus for canceling the frequency drift of a hydrogen maser as defined in claim 6, wherein The fitting of the preset linear model with the long-term component signal includes: Performing a filtering process on the long-term component signal to obtain a processed long-term component signal. Fitting a parameter of the preset linear model with the processed long-term component signal.
10. The apparatus for canceling the frequency drift of a hydrogen maser as defined in claim 6, wherein The training method of the signal prediction model includes: Obtaining a plurality of training samples and labels corresponding to the training samples, wherein each training sample is a frequency signal of the hydrogen atomic clock in a first time, and the label corresponding to the training sample is a frequency signal of the hydrogen atomic clock in a second time, the first time and the second time being consecutive and non-overlapping. Inputting each training sample into a preset signal prediction model to enable the signal prediction model to output a predicted value of the frequency signal of the training sample in the second time. Splicing the training sample and the predicted value of the frequency signal of the training sample in the second time to obtain a spliced signal, and calculating an Allan deviation of the spliced signal as a first loss term. Calculating a deviation between the predicted value of the frequency signal of the training sample in the second time and the label as a second loss term. Adding the first loss term and the second loss term to obtain a loss of the signal prediction model. Updating parameters of the signal prediction model to obtain a trained signal prediction model, with the objective of minimizing the loss of the signal prediction model.
Citation Information
Patent Citations
Atomic time calculation method and device based on hydrogen atomic clock drift prediction
CN114818247A
Atomic clock frequency anomaly detection method
CN120704101A
